Engagement rate is not a vanity metric by default. It becomes one when it is the only number you report, or when you compare it across accounts of different sizes. Used against your own median, it shows which posts work. Used alone, it hides whether anything paid off.
TLDR: Keep engagement rate to compare your own posts. Stop using it to compare accounts, to prove results to a client, or to set targets. Add four numbers beside it: shares and saves, link clicks, tracked conversions and inquiries.
Mydrop data, September 2026: across accounts connected to Mydrop, which manages 55,000+ posts, the median engagement rate per post was 3.1% on Instagram, 3.0% on TikTok and 1.1% on Facebook. The middle half of Instagram posts ran between 1.1% and 6.7%, so a 2% post is normal and a 7% post is a good day. Method and full tables in the Mydrop social media benchmarks.
Three ways engagement rate misleads you
1. The audience size trap. Rates fall as accounts grow and as they post more. In Mydrop's data, accounts posting under once a week had a median rate of 7.0%, while accounts posting 7 or more times a week had 0.7%. The correlation is partly about account size, so a "higher rate" can simply mean a smaller, quieter account.
2. All engagement counts the same. A like takes one tap. A share or a save means someone wants to come back or show a friend. A rate that adds them together hides the difference.
3. It can go down when your results go up. A link in the caption cost 7% of engagement in the same dataset, because clicking away is not an engagement. A post that sends 80 people to your site can look worse than a meme that sends none.
Four numbers to put next to it
| Metric | What it tells you | Where to find it |
|---|---|---|
| Shares and saves per post | Whether people want to pass it on or keep it | Platform insights, or Mydrop's per-platform metrics such as Pinterest saves |
| Link clicks per post | Whether a post moves people off the platform | Tagged links, link-in-bio clicks |
| Conversions from tagged links | Whether those people did the thing | Your analytics tool, with UTM tags on each link |
| Inquiries (DMs, emails, calls) | Whether the right people noticed | Inbox plus a "how did you hear about us" question |
You do not need all four. Pick the one closest to revenue and review it monthly. For a restaurant that is calls and bookings. For a SaaS it is trial signups from tagged links.
Set targets from your own median, not from a viral post
A target above the 75th percentile is a stretch goal, not a KPI. Take your last 30 posts, find your median engagement rate, and set the goal as "median plus a few points" for one format. If your Instagram median is 3% and your top quarter starts at 6%, a goal of 4% is real and a goal of 10% is a wish.
For the benchmark line, use the numbers in the block above. Half of Instagram posts land between 1.1% and 6.7%, so most of what you publish is already normal.
A three-question test for any metric on your dashboard
Before a number earns a place in your report, ask:
- Could it change what we post next week? If not, drop it.
- Does it measure the audience we want? Followers from the wrong country can lift reach and lower sales.
- Can we compare it to our own history? A metric with no baseline is a number without meaning.
Engagement rate passes all three for comparing your own posts. It fails the second and third when it is used across accounts.
What to tell a client who loves engagement rate
Do not argue with the number. Add one more to the report. Show engagement rate for the month, then show clicks and the one conversion that matters to them, side by side. If both rise, you have your story. If engagement rises and clicks do not, you have your next test.
Mydrop's performance dashboards show native metrics per platform, and white-label reports let you put engagement, clicks and saves into one monthly PDF without rebuilding it each time.
Next step: open your last report and circle every number that could not change a decision. Remove it, and put one conversion metric in its place.
Google review
Trustpilot review
Product Hunt comment
